The manufacture of cotton, wool, and synthetic fiber textiles is one of the world's largest industries. Moreover, some specific exposures within the industry are either known or suspected carcinogens. In spite of these considerations, existing knowledge of cancer risks to textile workers is based on fragmentary epidemiologic data. We are proposing an epidemiologic study in a cohort of roughly 267,000 women employees in the textile industry in Shanghai, China. The study cohort has been enumerated previously for a randomized trial of the efficacy of breast self exam, and is well characterized with respect to demographic, reproductive, and lifestyle factors, including cigarette smoking and alcohol use. We will focus on the following exposure/disease associations as primary hypotheses, all of which have been suggested but remain largely inconclusive in the literature: cotton and wool dusts and sinonasal cancer; formaldehyde and nasopharyngeal cancer; cotton dust and lung cancer; textile dyes and urinary bladder cancer; synthetic fibers and colon cancer. In a more exploratory mode, we will investigate textile exposures as potential etiologic factors for breast cancer. The study will consist of two related phases. The first phase will involve comparisons of site-specific cancer risks between the cohort and rates in the general population of Shanghai women during 1989-97. Incidence rates will be compared with city rates for the entire cohort and for the various manufacturing sectors (cotton, wool, synthetics, silk, dyeing, and finished apparel). The second, more in-depth analytic phase will be a case-cohort study nested within the cohort. The case groups will include incident, during 1989-97, cancers of the lung (expected number 1248), sinonasal passage (21), nasopharynx (100), bladder (81), colon (433), and breast (1196). A common referent subcohort (N=2496), will be selected as an age-stratified random sample from the study base of women textile workers. Historical exposure reconstruction will be performed for cotton, wool, silk, and synthetic fiber dusts, dyes, and formaldehyde to support dose-response estimation. The proposed study will be comprised of unquestionably the world's largest, most well characterized cohort of textile workers, and should therefore generate important information that is needed for cancer risk reduction strategies for women in China and elsewhere, including the United States.